ABSTRACT
The ecological environment in southern Xinjiang is fragile. Rainstorms and flood disasters caused by short-duration heavy rainfall often lead to huge losses in local agriculture and animal husbandry. In this study, five disaster elements causing rainstorms and flood disasters in southern Xinjiang are used to construct a disaster exponent that can comprehensively represent the severity of the disasters by using the ratio method and dimensionless linear combination method. Based on the probability distribution function, the county and annual disaster loss indices are classified into four grades, mild (grade 1), moderate (grade 2), severe (grade 3), and extremely severe (grade 4). Further, the spatial and temporal distribution of disasters and their causes are analyzed. Results show that the five disaster elements exhibited significant geographical variability, and areas with the most severe (grade 4) rainstorms and flood disasters occurred in the Aksu and Kashgar regions. The disasters appeared from March to October and were concentrated from May to July. The interannual variation of the disaster exponent from 1986 to 2019 showed a significant linear decreasing trend, by 0.4 per 10 a. The 12-h rainfall threshold that triggered rainstorms and flood disasters in southern Xinjiang was 9 mm, and the spatial distribution of the number of occurrences of 12-h precipitation that exceeded the threshold determined the spatial distribution of disaster severity. There is a significant positive correlation between the annual occurrence of 12-h precipitation exceeding the threshold (N 12) and the annual disaster exponent (Z K), and the decrease of N 12 leads to a decrease in the intensity of rainstorms and flood disasters in southern Xinjiang.
